Back

Alibaba

Qwen3.6 27B

ReasoningOpen weightsModels.devHugging Face
Intelligence
21.4
Rank 227 of 744 measured models
Coding
53.7
Rank 108 of 298 measured models
Output speed
55.6 t/s
Rank 301 of 403 measured models
Price/1M
$1.35
Rank 418 of 659 measured models
Context window
262K
Rank 458 of 1199 measured models
Cost per task
$0.621
Rank 112 of 193 measured models

Qwen3.6 27B (Reasoning) scores 21.4 on the Artificial Analysis Intelligence Index, ranking 227th of 744 measured models. The closest models on this index are G9v3-39A5B (21.8), KAT Coder Pro V2 (21.7), and A.X-K2 (21.5). At $1.35 per million tokens (3:1 input/output blended price), it is more expensive than KAT Coder Pro V2 ($0.525).

Artificial Analysis Indices
Reasoning
Intelligence21.4
Rank 227 of 744 measured models
Coding53.7
Rank 108 of 298 measured models
Agentic18.5
Rank 100 of 189 measured models
Reasoning levels
Each point is a reasoning level of this model: more reasoning often raises the score but costs more. Select a point to open that level.
Intelligence Indexmost efficient ↗Price / 1M tokens
Capability indexes
Performance in specific industries, measured by Artificial Analysis
Finance & Accounting21.3
Rank 152 of 218 measured models
Strategy & Ops16.9
Rank 150 of 218 measured models
Legal23.3
Rank 152 of 218 measured models
Engineering23.3
Rank 148 of 217 measured models
Economics31.4
Rank 146 of 225 measured models
Standard Benchmarks
GPQA Diamond84.2%
Rank 153 of 619 measured models
HLE23.1%
Rank 198 of 664 measured models
SciCode42.8%
Rank 157 of 231 measured models
IFBench67.6%
Rank 86 of 450 measured models
AA-LCR77.3%
Rank 130 of 575 measured models
TerminalBench Hard34.8%
Rank 83 of 432 measured models
Terminal-Bench 2.160.7%
Rank 102 of 244 measured models
Terminal-Bench v40.0%
Rank 151 of 218 measured models
Models at a similar level
The closest models on the Intelligence Index, one configuration per family.

Preferences and analytics

Rybbit: visits, performance, clicks, copies, forms and session replay (masked inputs), with your permission. Refusing does not limit the site.

Privacy

Allow navigation analytics?

With your permission, Rybbit measures visits, performance and interactions (clicks, copies, forms), including session replay, to improve BenchSift. Input values are masked in replays. Refusing does not limit the site.

Your choice is stored for 180 days. Change it at any time using “Privacy preferences” in the footer. Withdrawal reloads the page.

No choice stored: Rybbit is disabled.

Privacy